DATABASES Signal 205
pg_statviz 1.2 released with PostgreSQL 19 support and new features
↗
Illustration only Photo by Brecht Corbeel on Unsplash
pg_statviz 1.2 extends the extension and utility pair with PostgreSQL 19 compatibility, new wal_fpi_bytes metric capture, a blocking locks analysis module, and optional OpenAI AI integration while remaining lightweight and opt-in.
Engineers gain visibility into new WAL I/O metrics and lock contention without installing external agents, as the extension runs inside the database. The optional AI provider lets teams automate severity assessments while keeping the core tool dependency-free.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
pg_statviz 1.2 adds support for PostgreSQL 19 and captures the new wal_fpi_bytes counter from pg_stat_wal.
The release introduces a blocking locks analysis module that records blocked and blocking sessions by lock type using pg_blocking_pids().
It adds an OpenAI AI provider option, updates default AI models to claude-sonnet-5 and gemini-3.7-flash, and keeps AI features opt-in with zero-dependency base installs.
THE READ
What the cluster adds up to.
pg_statviz 1.2 adds support for PostgreSQL 19 and captures the new wal_fpi_bytes counter from pg_stat_wal. It also introduces a blocking locks analysis module that records the number of blocked and blocking sessions, along with a breakdown by lock type. Detection is built on pg_blocking_pids(), so even soft blocks are counted. The release has been tested against PostgreSQL 19 beta3 and across the whole PostgreSQL 13 to 19 range.
pg_statviz remains a lightweight extension and utility pair that lives inside the database, requiring no invasive agents or open external connections. The extension is plain SQL and PL/pgSQL and does not require additional modules to be loaded; the visualization utility is separate and can be run from anywhere. AI analysis is fully optional and opt-in; if the --ai flag is omitted, pg_statviz behaves exactly as before with no extra dependencies and no external calls. For AI features, the openai package can be installed via pip install pg_statviz[ai]; the base install stays zero-dependency.
When the AI provider is used, the tool needs network access to the OpenAI-compatible endpoint specified by OPENAI_BASE_URL and OPENAI_MODEL. If the --ai flag is not supplied, no external calls are made and the tool functions without any network dependency. The blocking locks analysis depends on the pg_blocking_pids() function, which is available in the supported PostgreSQL versions (13-19) but may not exist in earlier releases. pg_statviz does not aim to replace comprehensive monitoring platforms that collect system-level metrics or require external agents; it focuses on lightweight, database-resident statistics.
Written by elseif from the cluster below · checked for specifics the sources never containedTHE CLUSTER